A preprint is a signal, not a finished product or an independently confirmed result. We therefore track research papers separately and show the actual state of supporting evidence for each one.
1381
published research events
Latest work
A significant claim from a single source is published only after further confirmation.
The study compared five variants of an agent on 668 tasks. Memory with free-form notes increased the first-attempt success rate by 4.6 percentage points, while memory supplemented with reusable code increased it by 2.8 points. The difference between the two designs was not statistically significant overall.
The RoundtableReview framework uses six specialized AI reviewers to assess scientific papers. The authors report that, in two automated research systems, it improved paper quality compared with baseline methods and generated fewer comments.
According to a study using generative AI, if emissions continue to rise, once-in-a-century daily rainfall in New Zealand will intensify by an average of 9 % per degree of warming. AI generated more than 15 000 simulated years in six hours on four specialized chips.
The authors propose assessing inequalities based on the availability, quality and number of autonomous AI agents. They argue that delegating goals and direct competition between agents may create new power asymmetries. The available abstract presents a conceptual framework, not measured results.
The study introduces TrajectoryFlowNet, a network for reconstructing flow fields and long particle trajectories from sparse measurements. The authors validated it using numerical examples and experimental blood flow data. The available excerpt does not report numerical accuracy results.
An expert commentary describes a neural network that combines Lagrangian and Eulerian perspectives on flow. According to the available excerpt, it reconstructs physically accurate, high-resolution maps from sparse observations; the excerpt contains no specific measurements.
Irene Solaiman, public policy director at Hugging Face, highlighted the difficulty of reproducing and comparing AI evaluations. According to her, external evaluators need more consistent qualification requirements, and testing agents as a whole is both challenging and expensive.
The authors of the study Emerging Trends in AI Agent Evaluation stress that evaluating agents requires checking the entire process. According to the authors, evaluation credits alone for several agents in the ExploitBench test cost tens of thousands of dollars.
AI Snake Oil (Narayanan & Kapoor, Princeton) — a critical counterpoint to hype
Original source ↗
Researchers from MIT, Boston Children’s Hospital and Inria developed the E(3)-Pose model, which uses the symmetry of the fetal head to track its orientation during MRI. According to the authors, it is intended to help produce higher-quality 3D images; the available excerpt does not report numerical results.
The available excerpt links ten minutes of AI use to reduced persistence and poorer performance when subsequently solving tasks independently. The excerpt does not describe the method or the scope of the findings.
The study presents BRIA, an agent that isolates tool testing and incorporates only useful results. The authors tested it with 1 897 tools and three models. On MCP-Atlas, they report an average claim score of 89.77 compared with 75.40 for MCP-Zero.
MetaCurate creates filters for training trajectories using pairwise LLM evaluations. According to the authors, it achieved the highest average SWE-Verified score among the methods compared in subsequent experiments; the search process for the comparison method involving repeated training was 3.5 times as costly.
A study of four families of MoE models found that drops in F1 when limiting the number of experts are often caused by the response format. Keeping the full number of experts for about 4 % of tokens kept F1 within one point of the baseline in some comparisons. Time to first token fell by about 14 % for inputs with 11 thousand tokens.
A study of 42 288 predictions found that routing based on answer confidence helps with reading comprehension and topic classification, but does not deliver consistent improvements on knowledge and reasoning tasks. Differences between languages in forwarding requests to larger models reached 30.33 percentage points.
According to the study, a combination of initial LLM-generated proposals and subsequent optimization reached the target accuracy in 9 simulator runs, the LLM alone in 12 and the Nelder-Mead method in 17. However, the advantage of physics-based error descriptions disappeared after 15 evaluations.
The study combines search, evaluation by language models and Python execution to detect hallucinations in Bengali. Kaggle reports a macro-F1 of 0.956; however, an audit found overlap between the corpus and 38 evaluated inputs. The study does not measure the contribution of individual components under controlled conditions.
The study authors introduce HeRE, a dataset with 1.5 million automatically annotated examples. The method uses LLMs to clean up links between knowledge bases and texts. The model Gemma-4-31B-it achieved an F1 score of 0.844 on a held-out test set with two annotations.
The authors introduce BayesianPRM, which accounts for uncertainty in the evaluation model and the difference between training and deployment policies when evaluating reasoning steps. According to the authors, the method produced more accurate reward estimates and improved performance with test-time scaling in multimodal tests.
AI Radar monitors Czech and international sources every day, looking for changes that truly deserve attention.
MonitorsOfficial AI company blogs, specialist media, and research sources.
Selects and combinesFilters out information noise and combines articles about the same change into a single event.
Summarizes and explainsExplains significant events in English: what happened, why it matters and where the information comes from.
The result is a quick overview of what has actually changed in the AI world, rather than another stream of articles.
Use the CS/EN switch to read the same Radar in Czech or English. English content is published after its translation has been checked, so new and older items may appear later.
Everything you need to navigate the AI world
Today’s briefingAn overview of news from the past 24 hours and a main story when available. Beside them is Live · AI Flash, followed by research and links to other Radar sections. On mobile, the blocks appear one below another.
AI FlashAn ongoing feed of brief updates with an evidence status. Links lead to a Radar detail page when one is ready, otherwise to the original source. You can also find reset and outage histories here.
Practical applicationsWhat new tools and features can do, what you can try and what their actual impact could be.
Model selectionModel comparison by type of work, capabilities, price and speed.
Research and archiveA separate research overview, topic search and older events by date.
One event, everything that matters
Each row represents one event — not one article. At a glance, you can see its significance, credibility and main point.
Illustrative example, not a current news item.
Importance: ▮▮▮ majorOpenAIModels✓ 6
Agent mode is available to all paying users
Until now, the mode was available only on the highest plan; it is now available on all paid tiers without a waitlist.
▮▮▮ major · ▮▮ important · ▮ we're tracking = how significant the change is✓ 6 = six independent publishers, not the number of articles or feeds✓ official = a clear release, law or incident is substantiated by the relevant authority1 source = no independent confirmation yetbold = who is behind the changegray text = a brief summary of what happened
The detail page contains a fuller summary, its significance and original sources. Practical impact appears in the detail and the For individuals and For businesses views. An AI Flash item reaches the main selection only after it has been expanded and meets the publication rules.
The same news, two practical uses
We first summarize each event in the same way for everyone. Based on those same facts, we then explain what the change means for your own use and what it could mean for how a company operates.
For individualsWhat you can use or try, how the change can help you at work and what to watch out for.
For businessesWhat impact the change could have on processes, costs, risks and other business decisions.
Today’s briefing is the same for everyone. Pages
For individuals and For businesses
can be found in the main navigation — they select only events relevant to the given use case.